Theory Weekly Highlights for October 2000
October 27, 2000
The numerical pollution problem in the 2D version of the TWIST-R code was solved by implementing a new 5 point finite difference scheme. The new 2D code now converges correctly and accurately reproduces the 1D results for an explicitly 1D cylindrical equilibrium. Some minor numerical problems remain near the magnetic axis which result from the singular nature of the Jacobian at the axis, but these problems converge away with increased mesh resolution and it appears to be possible to eliminate them completely in principle by improving the accuracy of some key coefficients. Once these are resolved and benchmarking against the PEST-III code is completed, it is expected that TWIST-R will be ready for public use. The new code should prove reliable for the full range of Advanced Tokamak equilibria, including cases not previously possible with PEST-III, as well as for non-tokamak alternates.
October 20, 2000
Using a realistic quasilinear rf operator to self-consistently simulate ICRF minority heating in a Monte-Carlo drift orbit code, we found (1) co-current majority ion rotation with low field side resonance heating due to favorable scattering of “potato” co-passing orbits inward, (2) counter-current rotation with high field side resonance due to favorable detrapping of boundary layer particles into counter-passing orbits toward the magnetic-axis. These features appears to be robust in a steady-state rf driven system. Detailed study with finite wavenumber is in progress.
October 06, 2000
In collaboration with FARTECH, predictions of poloidal Mirnov signals utilizing instabilities computed by the GATO code were produced using the matched filter technique for two different DIII-D discharges. For the L Mode edge, Negative Central shear discharge #87009, the predicted signals for an ideal instability were compared with the signal from the final disruption in that discharge. In the second case, the analysis was performed for an ideal instability computed with no wall and compared to that from the observed resistive wall mode. In both cases, remarkably good agreement with the measured signals was obtained.
Disclaimer
These highlights are reports of research work in progress and are accordingly subject to change or modification